Impedance relaying through a transformer 1

[Mbrooke]

Ok, lets say you need to provide secondary short circuit and internal fault protection for a 115kv - 34.5kv D-Y transformer on a radial transmission line supplying a rural distribution substation. Would it be better to to use step distance relaying with an over reaching zone through the transformer or over current elements instead? Anything special to consider when adding the positive sequence impedance of the transformer to the positive sequence impedance of the transmission line or is it way more complicated?


The transformer in question is 25/35/40 MVA with an R of 0.01363 and an X of 0.43186



{Yes I know people will ask why not send a DTT signal to the supply station when differential kicks in, but I'd like to have redundant protection if the PLC fails}

 

[stevenal]

If you overreach with step distance, you will likely mis-coordinate with the 34.5 overcurrent elements. I find the transformer impedance is handy to help prevent zone 2 overreaching the transformer and mis-coordinating in this manner. Add coordinating time-overcurrent elements to provide the redundancy.

 

[Mbrooke]

Valid point. Would a time delay on zone 2 (the over reaching zone) take care of coordination 0of secondary feeder faults?

 

[davidbeach]

We use a distance zone that will reach through the transformer, but at 120 cycles it is pretty much a last ditch effort to clear an otherwise uncleared fault. Reaching through the transformer is like reaching several additional lines beyond the remote station, so it needs a bit of coordination time to make sure it isn't a line fault 3 lines away.

As for setting it, impedance is impedance. Put the transformer into your fault model and place a fault on the low-side bus. If the transformer is tapped from a line where one end is a much stronger source than the other, you may find that you need to let the strong end clear before the weak end can do anything. It can be a bit of a double-edged sword, look at a bunch of cases and see what seems to work and what might present an undesired level of risk.

 

[Mbrooke]

From your experience, would there be any disadvantage if I protect the transformer via over current only while only having a single 80% non delay zone 1 for the line itself?


FWIW the 115kv line is radial, but the source does vary in strength depending on generation dispatch.

 

[davidbeach]

Why overcurrent relays? The whole point of distance is that source strength ceases to matter and you reach approximately the same point regardless of source impedance.

For radial lines into transformers it becomes even easier. We try to set a zone 1 that covers the entire line but doesn't reach through any of the transformers. If one of the transformers faults the line and the transformer should trip together and the transformer should be gone when the line recloses. Then you can have a distance element that reaches through the transformers without the worry about how many additional lines it covers seeing as how there aren't any more lines beyond the far end. Those are also the lines where the ground distance can be set the most sensitively since there's no worry (in our system where all the transformers in question are Dyn1 type transformers) that block low-side ground faults from being seen by the ground distance elements on the high-side.

 

[stevenal]

Why not add a delayed zone 2 element at about 125% of your line. That way your line is fully covered, and you've allowed some time for transformer relaying to operate. Zone 2 will not reach past the transformer. Then apply time overcurrent to coordinate with downstream elements.

 

[Mbrooke]

Why overcurrent relays? The whole point of distance is that source strength ceases to matter and you reach approximately the same point regardless of source impedance.

Just a consideration, as 120 cycles for a zone 2 would be unacceptable in my case. Unless you are referring to a zone 3 on looped lines. I guess this delay is what made me confused.

For radial lines into transformers it becomes even easier. We try to set a zone 1 that covers the entire line but doesn't reach through any of the transformers. If one of the transformers faults the line and the transformer should trip together and the transformer should be gone when the line recloses. Then you can have a distance element that reaches through the transformers without the worry about how many additional lines it covers seeing as how there aren't any more lines beyond the far end. Those are also the lines where the ground distance can be set the most sensitively since there's no worry (in our system where all the transformers in question are Dyn1 type transformers) that block low-side ground faults from being seen by the ground distance elements on the high-side.

There are 34.5kv lines beyond the far end, so any fault on the 34.5kv system must coordinate with what ever reaches through the transformer. Ie, a 34.5kv bus fault or 34.5kv line fault right in front of the substation must be cleared by the 34.5kv breakers before the 115kv protection elements decide to trip.

 

[Mbrooke]

Why not add a delayed zone 2 element at about 125% of your line. That way your line is fully covered, and you've allowed some time for transformer relaying to operate. Zone 2 will not reach past the transformer. Then apply time overcurrent to coordinate with downstream elements.

But in this case I'd like to have zone 2 reach through the transformer and trip for faults right in front of the transformer.


The transformer does have differential, but it relies on PLC for DTT. If the PLC fails, I'd still like to have a last resort fail safe.

 

[cranky108]

I would say zone 1 to cover 80% of your transformer impedance, and use the over current to coordinate, and cover your mech damage curve of your transformer.

Or use zone 1 reach your transformer, and zone 2 for the 80%.

If you had communications, set up a POTT scheme.

 

[Mbrooke]

There is a PLC for DTT, but how will POTT enhance the protection? Remember this is a radial line.

 

[stevenal]

If you reach through the transformer with distance, you still have a coordination problem with trying to stack a definite time curve with the inverse curves of the 34.5 kv stuff. It's unlikely to work for all faults without significantly delaying zone 2 as Davidbeach suggests. The last resort failsafe I'm suggesting is from time-overcurrent relaying that is set to reach through the transformer and coordinate with the 34.5 stuff.

Since the transformer protection trips the remote line end, there is no harm in reaching into the transformer with zone 1 as Cranky suggested.

 

[Mbrooke]

If you reach through the transformer with distance, you still have a coordination problem with trying to stack a definite time curve with the inverse curves of the 34.5 kv stuff. It's unlikely to work for all faults without significantly delaying zone 2 as Davidbeach suggests. The last resort failsafe I'm suggesting is from time-overcurrent relaying that is set to reach through the transformer and coordinate with the 34.5 stuff.

But my question is, what if I set zone 1 to 125% of the line and Zone 2 at 70-80 cycles?

 

[davidbeach]

Yep, you can set it that way, but you'll need to look at the coordination. Look at your transmission line source conditions and using the weakest configuration that makes sense check the coordination margin between the distance element at its maximum reach and the 34.5kV overcurrents. You want the weakest system as it won't have much impact on the reach of the distance element but will reduce the fault current and slow down the overcurrent.

 

[Mbrooke]

That makes perfect sense then, thank you for you help

Last thing regarding the furthest reach, I just add 100% of the positive sequence line impedance + 100% of the transformer positive sequence impedance x 1.25 and I am set with my zone 2 distance value? I saw 80% mentioned, but in all honesty I am a bit clueless how that number works out.